Last month I stood in a room in Frankston (local business owners, chamber members, a hairdresser who’d been running her salon for 40 years, a plumber, an IT specialist, a handful of people who’d built something real in a community they clearly loved) and I asked them a question that sat in the air longer than I expected.

Not how do you compete with Amazon. Not what does AI mean for your future. Something simpler.

What is it, specifically, that you offer that can’t be replicated at scale?

The room took a moment.

Then the answers came: the personal welcome. Knowing someone’s name before they ring you back. Treating a client’s home like your own. The feeling, when you walk into a small business and someone actually sees you, that you’re not just a transaction.

I used a word for it. Ceremony.

Not the grand ceremonial sort. The small, repeatable kind. The cup of coffee made by someone who asked how your week was. The tradie who notices your garden on the way in and mentions it on the way out. The kind of interaction that makes you ring the same person next time, without shopping around, even when someone cheaper exists.

The businesses in that room that were holding their own weren’t doing it by getting better at convenience. They were going in a completely different direction.

That’s the signal I want to sit with today.

The Pattern Nobody Talks About

We’re in the middle of a productivity acceleration unlike anything most organisations have experienced. AI tools are compressing work that used to take days into hours. Research that required teams now takes one person and a good prompt. Writing, analysis, code, imagery, legal summaries, customer service responses: across every function, the rate of output has increased dramatically.

And the almost universal assumption is that this is unambiguously good. That getting better, faster, at more things, is what the moment requires.

Collective acceleration changes what “getting better” actually means.

When everyone in a sector gets better at the same thing at the same time, the improvement stops being an advantage. It becomes the baseline. And what was previously a competitive edge (faster turnaround, more polished output, more responsive service) becomes the minimum expectation. The ground moves. The bar rises. And the race continues, in the same direction, at a higher speed.

Research on scientific literature has been tracking this pattern for decades. A 2023 study published in Nature analysed tens of millions of scientific papers and patents from 1945 to 2010. The finding was striking: despite a massive increase in the number of researchers, the number of papers, and the sophistication of research tools, the measurable “disruption” of science (its tendency to create genuinely new directions) had declined significantly over that period. Science was getting better at consolidating what it already knew. It was producing less of what it didn’t know yet.

More researchers. More papers. More tools. Less disruption.

The productivity went up. The diversity went down.

Sound familiar?

What “Getting Better” Does to a Landscape

This is where the Ripple Effects framework becomes essential, not just as a strategic concept, but as a practical map for what’s happening around you right now.

A ripple effect isn’t just a consequence. It’s a cascade. The first thing that happens leads to the second, which leads to the third, and by the time you’re at the third order, you’re often somewhere the original decision never intended to go.

Here’s how the cascade looks when everyone improves in the same direction.

First order: AI and productivity tools make your organisation more efficient. This is real, measurable, and often significant. The time saved is genuine. The output is higher. You’re faster than you were six months ago.

Second order: Every competitor you have access to the same tools. And they’re getting faster too. Which means the efficiency gains you worked for are now distributed across the landscape. What was your advantage last year is table stakes today. You’re competing harder than ever, with the added pressure that the floor has risen and there’s nowhere to hide. The businesses that win purely on efficiency are usually the ones with the deepest pockets to sustain it. Not, historically, the small businesses in Frankston. Not the mid-market organisations. Not the teams that got fast but not deep.

Third order: The organisations that opted out of the same-direction race (not by ignoring the tools, but by asking a different question about what they’re actually for) are now positioned in a way their competitors can’t easily replicate. Not because they’re more efficient. Because they went in a different direction while everyone else accelerated in the same one.

This is the third-order consequence nobody plans for.

The 74 and the 20, Revisited

I wrote earlier this year about PwC research showing that 74% of AI’s measurable value flows to 20% of organisations.

That’s the first order finding. It’s significant. But the second-order consequence is the one that should be keeping strategy teams up at night.

The 80% of organisations not capturing the value aren’t failing to adopt AI. Many of them are adopting it aggressively: buying the tools, training the people, updating the processes. They’re getting better. Just in the same direction as everyone else in the 80%.

The 20% who are capturing the value aren’t just using AI more. They’re using it differently. They’ve made deliberate choices about what to automate and what not to. About where human judgment matters and where it doesn’t. About which direction their improvement should go.

That’s not a technology gap. It’s a strategic direction gap.

And strategy gaps compound over time in a way technology gaps don’t. A technology gap can close. You can buy the same tools, hire the same people, install the same systems. But if you’ve been optimising in one direction and the 20% have been moving in a different one, the difference between where you arrive isn’t one year of catching up. It’s a gap in capability, culture, and customer relationship that no tool release can close.

From Two Rooms

I want to share two observations from recent rooms I’ve been in. One I can name. One I can’t.

The named one is Frankston City Council’s Better Business event in June. The conversation I described at the start of this post. A hairdresser (40 years, a salon in Langwarrin) mentioned she’d started using ChatGPT six months earlier. Within weeks, she’d been found by a new customer through an AI search, not a directory, not a website. Someone had found her because she’d put enough of herself into the conversation with the AI that it started to know her. Not her services. Her.

She said something I haven’t stopped thinking about: “I just put in information about who I am and what I do. And it knows me now.”

That’s not automation. That’s the opposite of automation. That’s putting ceremony into every channel available, including the new ones.

Meanwhile, every other business in that room was worried about whether AI was going to make them irrelevant. She was already using AI to make herself more present. Different direction.

The second observation is from a group of senior executives in the construction and infrastructure sector (a group I’m not naming) who spent a morning working through a framework I use called HUMAND: a structured way of thinking about which decisions belong to humans, which to machines, and which to AI, and why.

When I work through this with leadership teams, the instinct is often to go in one direction: identify what can be automated and do it. And that’s a legitimate question to ask. But the more valuable question (and the one that seemed to land hardest in that room) is the other one. What do only humans do well here? What is it that, if we automate it, we lose something we can’t name but will definitely notice?

The relevance rating from that session was 10 out of 10. Not because I gave them efficiency answers. Because I gave them a different direction to look.

The Judgement Problem at Scale

I’ve written before about the distinction between information, knowledge, and wisdom, and about how we don’t have an information problem, we have a judgement problem.

That distinction is directly relevant here.

When everyone improves in the same direction, what they’re almost always improving is the information and knowledge layer. Faster access. Better synthesis. More accurate outputs. The things AI does well.

What they can’t improve at the same rate or in the same direction is the wisdom layer. The capacity to read a room. To notice what’s unspoken. To know when the technically correct answer is the wrong one for this person at this moment. To make a decision that will hold up not just to logic but to the people who have to live inside it.

The hairdresser’s customers aren’t coming back for 40 years because she delivers technically superior haircuts. They’re coming back because she provides ceremony. The kind of wisdom-layer experience that accumulates into trust, loyalty, and the absence of a need to shop around.

The construction executives who found the HUMAND conversation most useful weren’t the ones most focused on which tasks to automate. They were the ones who’d been asking, quietly, about what they were at risk of losing if they automated everything that looked automatable.

That’s wisdom-layer thinking. It doesn’t get better in the same direction as everyone else. It goes somewhere else entirely.

What the Narrowing Actually Looks Like

The science literature finding is worth staying with for a moment longer.

The researchers who studied decades of scientific papers weren’t saying the science got worse. It didn’t. The papers were better written, more rigorously conducted, more carefully cited. By every quality metric, science improved.

What narrowed was the diversity of what science was trying to discover. More researchers chasing more incremental gains in well-established fields. Fewer people working at the weird edges, asking the questions that don’t have an obvious methodology yet.

The tools made it easier to do the established work better. They made it harder, not easier, to break away from the direction the tools were already good at.

This is not a metaphor for what’s happening in organisations. It’s a direct parallel.

AI tools are excellent at the established work. They’re remarkable at incremental improvement within known parameters. The businesses and organisations that will be hurt aren’t the ones who adopt them too slowly. They’re the ones who adopt them completely, without first asking what those tools are optimising toward, and whether that direction is theirs.

Three Questions to Take Into Your Next Strategy Session

None of this is an argument against efficiency. I’m not suggesting local businesses avoid AI, or that organisations shouldn’t improve. The tools are real. The productivity gains are real.

But the strategic question isn’t whether to get better. It’s what direction better should go.

Three questions worth putting on the table.

First: In what direction is your sector getting better, and is that the direction you actually want?

This isn’t rhetorical. Most sectors right now are getting better at speed, volume, and automation. Some of those improvements are genuinely useful. But if every competitor is making the same improvements, the improvements aren’t building you an advantage. They’re keeping you in the game.

Second: What would you lose if you automated everything that looks automatable?

Not what tasks would disappear. What would disappear with them. The junior staff who build capability by doing the difficult work. The client relationships that run through the person who does the work, not just the account. The organisational knowledge embedded in the process, not just the output.

Second and third order consequences don’t show up in the efficiency calculation. They show up later, when it’s harder to see where they came from.

Third: What’s your “ceremony”?

What is it that you offer, in every channel, every interaction, every touchpoint, that can’t be replicated at scale? That doesn’t get better at the same rate as the AI-assisted equivalent? That your customers and clients come back for, without needing to consciously choose it each time?

If you can answer that question clearly, you know which direction to go.

If you can’t, that’s the more urgent strategic problem than any technology question on your current agenda.

The Third-Order Opportunity

The third order consequence of everyone improving in the same direction is actually an opportunity, but only for the organisations that see it coming.

When the efficiency landscape flattens (when everyone is roughly as fast, as capable, and as well-equipped as everyone else) the premium goes to difference. Not to the organisations that are marginally more efficient. To the ones who went somewhere else while everyone else was racing to the same destination.

The Frankston hairdresser didn’t compete with the chains on price or convenience. She competed on ceremony, trust, and 40 years of knowing her customers. AI didn’t threaten that. She used AI to extend it.

The construction executives who found the HUMAND conversation most relevant weren’t asking how to become more efficient. They were asking what only they could do, and protecting it.

That’s the third-order move. Not faster. Different.

I’ve been watching this pattern across industries and organisations for more than 30 years. The organisations that create durable advantage almost never do it by getting better at what everyone else is already improving. They do it by noticing (often before the evidence is overwhelming) that the direction everyone is heading is creating an opening somewhere else.

The opening right now is not small.

When efficiency becomes the baseline, ceremony becomes the premium. When speed is table stakes, judgment is the differentiator. When AI can do what everyone thought required expertise, the humans who offer wisdom become rare.

Not because they resisted the tools. Because they chose a different direction.

Choose Forward.

The question to carry out of this isn’t “how do we get better at AI?”

It’s: in what direction are we getting better, and is that direction ours?

Choosing forward doesn’t mean choosing faster. Sometimes it means choosing sideways. Choosing toward the thing that makes you irreplaceable in a landscape where everything that can be replicated eventually will be.

The second and third order consequences of everyone improving in the same direction are already here, moving through sectors, compressing margins, raising baselines. Most organisations won’t feel them until they’re fully arrived.

The ones that will feel them least are already in the room, asking the different question.

Choose Forward.

Frequently Asked Questions